CLASSIFICATION BASED ON FERMENTATION REACTIONS 569 



conditions, while giving typical a-haemolysis with green coloration when cultivated 

 aerobically. 



Summarizing the observations recorded above we may say that a study of the 

 appearance of the colonies on blood agar plates, together with a test for the produc- 

 tion of a filtrable haemolysin, enables us to divide streptococci into three main 

 categories. 



(1) Hcemolytic streptococci. — These produce /9-hsBmolysis on blood agar plates. 

 They may be differentiated into two sub-groups : (a) those that produce a filtrable 

 haemolysin, and (6) those that do not. Among strains of Str. pyogenes two variants 

 are known. One produces a-haemolytic colonies aerobically but j^-haemolytic 

 colonies anaerobically ; it forms a soluble hsemolysin in broth aerobically. The 

 other forms completely non-hsemolytic colonies, but produces a soluble hsemolysin 

 of the type. 



(2) Streptococci giving cn-hcBmolysis. 



(3) Streptococci that have no action on blood media under the usual conditions of 

 testing. 



Classification Based on Fermentation Reactions. — ^The capacity of different 

 species or types of streptococci to ferment different substrates played a very large 

 part in the earlier attempts to separate this group of bacteria into its natural 

 elements. The classification of Andrewes and Horder (1906), based largely on the 

 earlier studies of Gordon (1902-03, 1903-04, 1905), and the more extensive and 

 detailed classification proposed by Holman (1916) (see also Floyd and Wolbach 

 1914, Lyall 1914, Broadhurst 1915), depended almost wholly on a selected series 

 of fermentation tests. It is hardly necessary to-day to set out these classifications 

 in their original form ; since few would now accept them as affording an adequate 

 basis for the differentiation of named species or types. This does not of course mean 

 that these fermentation tests are of no value in classifying the streptococci. The 

 reverse is the case. In more recent years, however, the tendency has been to 

 employ fermentation reactions as ancillary, rather than as primary differential 

 criteria. When a particular substrate has been found to be of value in differentiat- 

 ing between certain related species or types, it has been used for this purpose, but 

 has not necessarily been employed in the differentiation of other species within 

 the genus. It will therefore be more convenient to consider the exact systematic 

 significance of these fermentation tests after we have discussed certain other 

 differential criteria, and, in particular, the results obtained by antigenic analysis ; 

 but it will be useful to summarize here the observations that have been made on 

 the correlation between enzymic activities and natiiral habitat. 



The early observations that streptococci isolated from pathogenic lesions in man usually 

 ferment lactose and salicin, seldom, if ever, mannitol, raffinose or inulin, and give acid 

 without clot in milk has been amply confirmed. The value of inulin fermentation as a 

 differential test for the identification of the pneumococcus has also been firmly established. 

 It is clear that this species almost always ferments this fructosan, while most other strepto- 

 cocci fail to do so. There has been a tendency to exclude from the species Str. pneumonice 

 any strain that fails to ferment inulin ; but it is doubtful, as Berger and Silberstein (1926) 

 point out, whether this test can be applied with such complete rigidity. As regards other 

 substrates, the pneumococcus ferments lactose and usually raffinose, but not salicin or 

 mannitol. Milk is acidified and frequently clotted. 



There is some measure of agreement that the streptococcus commonly found in the 

 human mouth, which is of the a-haemolytic type, ferments lactose, raffinose and saUcin, 

 but not mannitol, and usually forms a clot in mUk. In addition, it has been pointed 



